Precision Agriculture has been experiencing an essential growth in the implementation of industrial internet of things based applications. The proposed evaluation framework uses a hybrid decision-making model for technology selection. The structure combines two extensively used methods, Analytic hierarchy Process (AHP) and Grey Relational Analysis (GRA). The evaluation process of data acquisition system features is two-fold. First, AHP is used to assign importance to the criteria. Next, GRA is used to assess the alternatives concerning the criteria. Finally, we obtain the final grey relational coeficients for each alternative and chose the most suitable one.
plants, categorize them and predict yield and crop load [7], the collection of large data sets for a single corn plant [8], enhance energy eficiency [9], remote sensing using drones [10], climate change monitoring [11], detection of infections and crop classification [12], smart crop-field monitoring and automation of irrigation systems [13], the use of IoT devices for sensing the agricultural data and store data into the Cloud [14], the use of Data Mining Techniques and IOT to improve the crop yield [15], nutrient management for livestock [16, 17] and pollution-free cultivation systems among many others.
Analytic hierarchy process is a decision-making tool whose main objective is to interpret the intangible judgments of people as tacit preferences. It does so by decomposing the problem into several hierarchical levels. The first level is the objective or goal, next, the comparison criteria and sub-criteria, and at last, the alternatives [18]. Using AHP brings advantages such as handling uncertain information lowering the experience of decision-makers. The AHP technique can be applied in many fields. Among them, we can highlight project selection, prioritization of environmental impacts[19], mining method selection [? 20], budget allocation, medical decision-making, transportation [21], manufacturing an supply chain [22, 23, 24] among many others.
Grey relation analysis is applied in multiple fields. Yeh [25] proposed an early combination
with multi-criteria decision-making techniques in order to evaluate weapons systems. Later,
other developments arise combining GRA with AHP in diverse areas including the selection of
marketing networks [26], the impact analysis of damage in natural disasters [27] among others.
A formal description of GRA and AHP is given by [28]; later Song and Jamalipour[29, 30, 31]
extend the use of the AHP-GRA combination in the selection of networks in wireless
communications, the same year Li also applies this combination in the selection of materials [32] and
data transmission on demand. Zhang also implemented this combination in the evaluation of
knowledge management tools [
For solving the Decision Making problem, we propose the use of AHP, as an accepted and used
often in problems that include subjective judgments of people. According to Saaty [
⋮ ⎣⎢ ′ 1 ′ 2 1′2 2′2
⋮ ⋯
⋮ ⎥ ⋯ ′ ⎥ ⎦
Where w corresponds to the column vector of the relative weights obtained by making the average of each line of the normalized comparison matrix. tion of the original comparison matrix with the column vector of relative weights.
The value of max is obtained by adding the column vector corresponding to the multiplicaBecause comparisons are made subjectively, a consistency index is required to measure the consistency of the person making the ratings. The consistency index and the consistency ratio CR are calculated as follows:
= max max = ∑ = max − − 1 =
Where the RI inconsistency ratio is a comparison constant that depends on the size of the paired comparison matrix for sizes of n = 9 (our criteria x criteria matrix) RI = 1.45
comparable, the pressed as ′ = ( ′1, ′2, … , ′ ), where ′ means the ℎ
comparative factors can be expressed as:
Grey relational space theory introduced by Deng Julong [
Step 1. Generate the comparative factors. A set with components of criteria can be excriteria of ′ . If all criteria are (1) (2) (3) (4) is defined as follows (6): • In case of ‘smaller-is-better’ data treatment ( ), can be normalized into ∗( ) The formula Step 3. Compute the distance Δ0 ( ) between the reference series and the comparative series. = 0.5. risk factor , which is given as follows: where Δmin = min min (Δ ) ,
Δmax = max max (Δ ) , and is an identifier
afecting the relative value of risk without changing the priority. According to [
Step 5. Calculate the degree of grey relation based upon the and the group weights of Step 2. Determine the reference series and normalize all data sets. Degree of relation can represent the relation of two series; therefore, an objective series called reference series shall the following types: ‘larger-is-better’ or ‘smaller-is-better’. be established and expressed as 0 = ( 01, 02, … , 0 ). Data sets can be treated using one of is defined as follows (5): • In case of ‘larger-is-better’ data treatment ( ), can be normalized into ∗( ) The formula (5) (6) (7) (8) (9) where Δ Step 4. Ca=lcu‖‖ la0te−the ‖‖g.rey relational coeficient as follows: = Δmin + Δmax Δ + Δmax
, = 1, 2, … , ; = 1, 2, … ,
In this section we show how the use of a model that combines AHP-GRA can help to manage qualitative and quantitative information in the selection of technology for data acquisition for the beekeeping sector. Beekeeping has particular characteristics necessary for proper performance of its hives, such as location, temperature, humidity, population control of the swarms, among others. The model was developed for a case in Colombia, located in the rural area of Carmen de Carupa. Which has a relative humidity of 86% and an average temperature of 10 ∗( ) =
( ) − min ( ) max ( ) − min ( ) ∗( ) =
max ( ) − ( ) max ( ) − min ( )
⋮ ⎥ degrees Celsius. Next we will describe the Smart farming technologies, then we will describe the alternatives and the criteria of the model, finally we will apply the AHP-GRA combination in order to obtain the best alternative for the selection of data acquisition technology.
Smart farming technologies are classified into three principal classes that incorporate entire operation of Precision Agriculture: • Data acquisition technologies: this class includes all surveying, mapping, navigation and sensing technologies. • Data analysis and evaluation technologies: this class includes from computer-based decision models to complex farm management and information systems.
• Precision application technologies: this class contains all application technologies. In order to evaluate data acquisition technologies. First, we determine alternatives and criteria from an extensive literature survey, next we apply AHP in order to obtain the weights for all criteria. At last, we apply grey relational analysis to rank all alternatives and select the most suitable one. An schema of the evaluation process can be seen in Fig. 1 as follows:
• Global Navigation Satellite Systems: is the conventional name for satellite navigation systems that contribute independent geo-spatial positioning with global coverage. • LiDAR Sensors: Light Detection and Ranging are instruments that measure the distance from the target by laser.
• Thermal Cameras Systems: Thermal cameras have the ability to generate images related to the ambient temperature.
Many factors need to be considered in the selection of perception sensors and other technologies in agriculture, which are described below: • Easy Maintenance: Reliability is essential for perception-based applications. Sensor performance will quickly degrade when exposed to harsh agricultural environments, and frequent sensor maintenance is required. The ease of sensor maintenance is another consideration in selecting a perception sensor. • Low Cost: The cost is one of the most important factors that influence, since in some cases the farmer cannot have access to high-cost elements that allow him to follow up on bees or any type of food that he sows on his farm. • Easy to Use: The simple and fast use of the measuring instruments used in Precision Agriculture is essential when acquiring them, since some farmers do not know how to read or write, therefore a friendly technology is needed for them.
From the literature review, the alternatives and criteria most related to our case study are established and we build the hierarchy that can be seen in the figure 2 as follows.
AHP is used to determine the relative weight of each criterion. using the scale from 1 to 9 with diferent levels of importance proposed by saaty. The priorities are calculated by calculating the dominant eigenvector that belongs to the matrix of pairwise comparisons for the criteria. The consistency ratio of the matrix of even comparisons corresponding to the criteria is CR=0.06239. Te weights of all criteria can be seen in Table 2 as follows.
The qualitative and quantitative evaluations are added in the table below, the criteria = {1, 2} are qualitative. On the other hand, the criterion = 5 is quantitative and corresponds to the price in dollars of each item.
Step 1. Generate the referential series and the compared series. Step 2. Then the referential series of 0 = {6, 8, 100}, and the compared series of 1 = {2, 4, 100}, 2 = {6, 5, 300}, and 3 = {6, 8, 150}.
Step 3. Calculate the distance between the referential series and the compared series. Γ01(Lidar) = 0.387303333, Γ02(Thermal − cameras) = 0.832982667, Γ02(GPS) = 0.973003333 The priorities of the three potential technologies (in order with their grey relational grades) is GPS > Thermal cameras > Lidars.
In this article, the combined AHP and GRA procedure gives a solution, including quantitative and qualitative information. This coordinated approach has selected the technology based on the proposed evaluation model. The benefit with this approach is that it is quite adapted to deal with problems that involve qualitative and quantitative values, and enables the evaluation based on limited information.
Based on the result obtained in the model, we can conclude that the best technology to acquire data in apiculture is GPS. It will help to obtain a useful mapping of the terrain with an adequate angle and temperature for giving the right conditions to the development of bees population. Computers and Electronics in Agriculture 155 (2018) 103–120. doi:10.1016/j.compag. 2018.10.001. [5] N. Liu, R. Zhao, L. Qiao, Y. Zhang, M. Li, H. Sun, Z. Xing, X. Wang, Growth stages classiifcation of potato crop based on analysis of spectral response and variables optimization, Sensors (Switzerland) 20 (2020) 1–20. doi:10.3390/s20143995. [6] P. Liu, H. Yang, Three-way decisions with single-valued neutrosophic decision theory rough sets based on grey relational analysis, Mathematical Problems in Engineering 2019 (2019). [7] Y. Ampatzidis, V. Partel, UAV-based high throughput phenotyping in specialty crops utilizing artificial intelligence, in: 2019 ASABE Annual International Meeting, American Society of Agricultural and Biological Engineers, 2019. doi:10.13031/aim.201900293. [8] I. I. I. Colley R., J. Fulton, N. Douridas, K. Port, Finding value in digital agriculture technologies: A 2017 case study, in: ASABE 2018 Annual International Meeting, American Society of Agricultural and Biological Engineers, 2018. doi:10.13031/aim.201801555. [9] A. Rajput, V. B. Kumaravelu, Fuzzy logic–based distributed clustering protocol to improve energy eficiency and stability of wireless smart sensor networks for farmland monitoring systems, International Journal of Communication Systems 33 (2020). doi:10.1002/dac. 4239. [10] B. D. Gossen, M. R. McDonald, New technologies could enhance natural biological control and disease management and reduce reliance on synthetic pesticides, Canadian Journal of Plant Pathology 42 (2020) 30–40. doi:10.1080/07060661.2019.1697370. [11] S. Mulla, S. K. Singh, K. K. Singh, B. Praveen, Climate change and agriculture: A review of crop models, Springer Singapore, 2019. doi:10.1007/978-981-13-9570-3_15. [12] M. A. Khan, T. Akram, M. Sharif, M. Awais, K. Javed, H. Ali, T. Saba, CCDF: Automatic system for segmentation and recognition of fruit crops diseases based on correlation coeficient and deep CNN features, Computers and Electronics in Agriculture 155 (2018) 220–236. doi:10.1016/j.compag.2018.10.013. [13] R. Nageswara Rao, B. Sridhar, IoT based smart crop-field monitoring and automation irrigation system, in: Proceedings of the 2nd International Conference on Inventive Systems and Control, ICISC 2018, Institute of Electrical and Electronics Engineers Inc., 2018, pp. 478–483. doi:10.1109/ICISC.2018.8399118. [14] S. Rajeswari, K. Suthendran, K. Rajakumar, A smart agricultural model by integrating IoT, mobile and cloud-based big data analytics, in: Proceedings of 2017 International Conference on Intelligent Computing and Control, I2C2 2017, volume 2018-Janua, Institute of Electrical and Electronics Engineers Inc., 2018, pp. 1–5. doi:10.1109/I2C2. 2017.8321902. [15] N. Sneha, K. V. Sushma, S. S. Muzumdar, Precision Agriculture using Data Mining Techniques and IOT, in: 1st IEEE International Conference on Advances in Information Technology, ICAIT 2019 - Proceedings, Institute of Electrical and Electronics Engineers Inc., 2019, pp. 376–381. doi:10.1109/ICAIT47043.2019.8987333. [16] L. C. Campbell, Beneficial impact of precision nutrient management on the environment and future needs, Communications in Soil Science and Plant Analysis 25 (1994) 889–908. doi:10.1080/00103629409369087. [17] G. Laible, J. Wei, S. Wagner, Improving livestock for agriculture - technological progress from random transgenesis to precision genome editing heralds a new era, Biotechnology Journal 10 (2015) 109–120. doi:10.1002/biot.201400193. [18] T. L. Saaty, What is the analytic hierarchy process?, in: Mathematical models for decision support, Springer, 1988, pp. 109–121. [19] J. I. Romero Gelvez, M. Garcia-Melon, Influence analysis in consensus search—a multi criteria group decision making approach in environmental management, International Journal of Information Technology & Decision Making 15 (2016) 791–813. [20] J. Gelvez, F. C. Aldana, G. Sepúlveda, Mining method selection methodology by multiple criteria decision analysis—case study in colombian coal mining, in: International Symposium of the Analytic Hierarchy Process, Washington DC, USA, 2014. [21] S. Duleba, S. Moslem, Examining pareto optimality in analytic hierarchy process on real data: An application in public transport service development, Expert Systems with Applications 116 (2019) 21–30. [22] R. Septifani, I. Santoso, B. Rodhiyah, Risk mitigation strategy of rice seed supply chains using fuzzy-fmea and fuzzy-ahp (case study: Pt. xyz), in: IOP Conf. Series: Earth and Environmental Science, volume 230, 2019, p. 012059. [23] P. L. Singh, R. Sindhwani, N. K. Dua, A. Jamwal, A. Aggarwal, A. Iqbal, N. Gautam, Evaluation of common barriers to the combined lean-green-agile manufacturing system by two-way assessment method, in: Advances in Industrial and Production Engineering, Springer, 2019, pp. 653–672. [24] L. A. Ocampo, Applying fuzzy ahp–topsis technique in identifying the content strategy of sustainable manufacturing for food production, Environment, Development and Sustainability 21 (2019) 2225–2251. [25] M.-F. Yeh, H.-C. Lu, Evaluating weapon systems based on grey relational analysis and fuzzy arithmetic operations, Journal of the Chinese Institute of Engineers, Transactions of the Chinese Institute of Engineers, Series A/Chung-kuo Kung Ch’eng Hsuch K’an 23 (2000) 211–221. [26] C.-T. Lin, P.-F. Hsu, Selection of internet advertising networks using an analytic hierarchy process and grey relational analysis, International Journal of Information and Management Sciences 14 (2003) 1–16. [27] S.-P. Cheng, R.-Y. Wang, Analyzing hazard potential of typhoon damage by applying grey analytic hierarchy process, Natural Hazards 33 (2004) 77–103. [28] C.-T. Lin, S.-N. Hwang, C.-H. Chan, Grey number for ahp model: An application of grey relational analysis, volume 1, 2004, pp. 226–230. [29] Q. Song, A. Jamalipour, Quality of service provisioning in wireless lan/umts integrated systems using analytic hierarchy process and grey relational analysis, 2004, pp. 220–224. [30] Q. Song, A. Jamalipour, Network selection in an integrated wireless lan and umts environment using mathematical modeling and computing techniques, IEEE Wireless Communications 12 (2005) 42–48. [31] Q. Song, A. Jamalipour, An adaptive quality-of-service network selection mechanism for heterogeneous mobile networks, Wireless Communications and Mobile Computing 5 (2005) 697–708. [32] H. Cao, F. Liu, C. Li, C. Liu, An integrated method for product material selection considering environmental factors and a case study, Materials Science Forum 532-533 (2006)